Background Hospitalisation for acute exacerbations of COPD is associated with high risk of readmission. However, no tool has been validated to stratify patients at discharge for risk of readmission.

Aim To evaluate the ability of the 4 m gait speed (4MGS), a surrogate marker of frailty, to predict risk of future readmission in hospitalised patients with an acute exacerbation of COPD (AECOPD).

Methods 213 patients hospitalised with an AECOPD were recruited prospectively. 4MGS was measured on day of discharge. Logistic regression models were used to assess the association between 4MGS and readmission at 90 days after discharge.

Results Baseline characteristics of the cohort: 52% men; mean age 72 years; median FEV₁ 35%predicted. Mean (SD) 4MGS at hospital discharge was 0.61 (0.26) ms⁻¹. Significant increased rates of all-cause readmission at 90 days were seen across quartiles of decreasing 4MGS (Q4 fastest: 11.5%; Q3: 20.4%; Q2: 30.2%; Q1 slowest: 48.2%; p_trend<0.001). Compared with Q4, those in the slowest 4MGS quartile had unadjusted ORs (95% CIs) for 90-day readmission of 7.12 (2.61 to 19.44) for the whole cohort and 11.56 (3.08 to 43.35) in those aged 65 or over. A multivariate model incorporating 4MGS, Charlson Index, hospital admission in past year, FEV₁%predicted and number of exacerbations in past year in those aged 65 or over predicted 90-day readmission with a C-statistic of 0.86.

Conclusions The 4MGS, a surrogate marker of physical frailty, independently predicts the risk of readmission in older patients hospitalised for acute exacerbation of COPD.

Trial registration number NCT01507415.